42 Estimating Rotor Suspension Parameters from Runout Data 377 Fig. 42.5 Runout in polar coordinates—trial run Fig. 42.6 Slow-roll–corrected runout magnitudes—reference run A slow-roll correction is applied to the probe data. The average of the first five data points for each probe is used as the slow- roll correction. Figures 42.6, 42.7, 42.8, and 42.9 show the runout data after performing the slow-roll correction. Note that after the slow-roll correction, the runout data at low frequencies have a magnitude near 0 for the magnitude plots or begin near the origin for the polar plots. After performing the slow-roll correction, the trial-run runout data are subtracted from reference run data. The resulting runout difference is the runout that would be expected for a perfectly balanced rotor with only the trial weight applied. Figures 42.10 and 42.11 show the difference between the slow-roll–corrected runouts for each probe. In these plots, the runout is limited to values above 42% of the first rigid-body critical speed. This limitation is imposed to avoid calculating suspension parameter values at frequencies where the left suspension dynamic characteristics significantly affect rotor runout. The simple rotor suspension model used here is a complex, frequency-dependent spring with only direct stiffness and damping. The concepts of effective mass and damping are used to express the frequency-dependent stiffness, Z(ω) of the suspension: Z(ω) =M(ω)ω 2 +iC(ω)ω (42.1)
RkJQdWJsaXNoZXIy MTMzNzEzMQ==